The Impact of AI on SEO: A Clear-Eyed Look

AI changed SEO in three concrete ways. Writing a page now costs almost nothing. Search results now include generated answers that cite a handful of sources. And the reports you used to judge success no longer cover every surface where people find you.

What did not change is the entry ticket. Google states that to appear as a supporting link in AI Overviews or AI Mode, a page "must be indexed and eligible to be shown in Google Search with a snippet", and that "there are no additional technical requirements." Other engines work the same way in principle: they can only cite what their crawler fetched and their index holds.

It helps to separate four stages:

• Parametric: what a model knows without searching. It is frozen at a training cutoff. Nobody changes it this quarter.
• Access: whether a crawler can fetch and store your page. You control most of this.
• Retrieval: whether your page is pulled in for a given question. The engine decides, partly based on relevance.
• Generation: which retrieved pages the model uses and cites. The model decides.

Most of what you can fix lives in the access stage. For how generated answers sit alongside the classic results, see how AI search features affect organic results.

AI lets anyone publish thousands of pages in a day. Google's position is clear on both sides of that. Its guidance says "appropriate use of AI or automation is not against our guidelines." Its spam policy defines scaled content abuse as "when many pages are generated for the primary purpose of manipulating search rankings and not helping users."

OtterlyAI tested the volume shortcut in 2026 with two new domains, each carrying roughly 1,000 AI-generated posts. Otterly reports that Google removed one site after about 14 days and the other after about 90, with no manual action shown in Search Console for either. The site with a fictional expert author box went first.

The test has limits. It ran one site per arm, had no human-written control, and was run by a vendor that sells AI visibility tracking. It does not prove author boxes hurt. It does show that the removal can arrive silently.

What this means for you:

• Use AI to draft, research and edit. Do not use it to multiply thin pages.
• Put specific, checkable facts on the page: a price, a number, a one-line definition. That gives an engine a sentence it can quote.
• Check every fact before publishing. AI drafts get details wrong, and those errors can end up repeated in AI answers.

Our guide to SEO copywriting with AI covers a workflow for drafts that stay specific and correct.

Search results now carry generated answers above or instead of the familiar list of links. On Google there are two separate surfaces. Google says AI Overviews and AI Mode "may use different models and techniques" and can return different links, so treat them as two channels (see AI Overviews vs AI Mode). AI Overviews also do not trigger on every query.

Ranking well still matters, but it is not a guarantee of a citation. Ahrefs found in March 2026 that 37.9% of URLs cited in AI Overviews also appeared in the top 10 results. The rest came from deeper positions. If you used to win the answer box, featured snippets vs AI Overviews explains what is different now.

Outside Google, each engine has its own index and its own rules, as of October 2026:

• ChatGPT search depends on OpenAI's OAI-SearchBot. OpenAI says sites opted out of it "will not be shown in ChatGPT search answers."
• Microsoft Copilot grounds its answers on the Bing index. In Otterly's test, 2,718 of the 3,934 citations tracked for one site came from Copilot, after Google had already removed that site.
• Perplexity and Claude run their own search crawlers, PerplexityBot and Claude-SearchBot.

Every engine decides per question whether to search at all. When it does not search, the answer comes from training data you cannot influence this quarter.

This is where AI hurt SEO the most, and nobody has fixed it. Google's generative AI report in Search Console shows impressions for AI Overviews and AI Mode. As of October 2026 it shows no clicks, no click-through rate and no position. AI Overviews also send no referral marker your analytics can read.

ChatGPT adds utm_source=chatgpt.com to the links it sends, but free-tier and mobile-app clicks often arrive untagged. Bing Webmaster Tools has an AI Performance report with citations and the grounding queries Copilot actually ran, but it has no API.

So the plain answer is this: the effect of AI answers on your organic traffic is not measured by any published source, and you cannot measure it fully from first-party data either. Any single "AI visibility score" averages answers given with and without live search. Prompt trackers sample too. In Otterly's test, Microsoft's own report counted 42,817 citations for one site, against 3,934 logged by Otterly's prompt-based tracker over windows that may not match.

What to do instead:

• Track each engine separately. One engine can cite you heavily while another never does.
• Record presence and accuracy (are you named, and are the facts right), not a rank. No engine publishes position data to check a rank against.
• Read Search Console and Bing Webmaster Tools side by side.

The pattern is simple. The controls you hold mostly remove blocks. None of them adds a guaranteed citation. When engines cite other sites instead of yours, being named on those sites is associated with appearing in the answer, though nobody has measured a reliable effect.

Work from the gates outward. A page that a crawler cannot reach gains nothing from better copy.

1. Read your robots.txt per bot. GPTBot controls training use only. OAI-SearchBot controls ChatGPT search. OpenAI says each setting "is independent of the others." A setup that opts out of training but stays in search looks like this:

User-agent: GPTBot
Disallow: /

User-agent: OAI-SearchBot
Allow: /

2. Check the firewall, not just the file. A CDN or bot management rule can return 403 to a crawler that robots.txt allows. From 15 September 2026, Cloudflare blocks Training and Agent crawlers by default on ad-displaying pages for new domains and existing free-tier customers. The AI Crawler Checker requests a page as each AI crawler and shows which ones get turned away.
3. Check rendering. Vercel found ChatGPT and Claude crawlers fetch JavaScript files but do not execute them. Prices, product names and key claims should be in the raw HTML.
4. Audit snippet controls. A template-wide nosnippet or data-nosnippet keeps content out of AI answers.
5. Confirm indexing engine by engine. Check Google Search Console and Bing Webmaster Tools separately. In Search Console, also check the generative AI inclusion control, which child properties can inherit from a parent.
6. State your facts once, clearly and consistently. Pricing, founding year and product names should match on every page.
7. See what AI says about you. The free AI SEO checker tests one page. The free checker on this site shows whether AI names you for up to three buyer questions.

• A blanket AI block. Disallowing every AI user agent also removes you from ChatGPT search, Perplexity and Claude search.
• Expecting schema or llms.txt to earn citations. Google says structured data isn't required for generative AI search and that Google Search ignores llms.txt. Ahrefs' matched test found AI Overviews citations fell 4.6% after pages added schema. Removing existing markup is not supported either: it is simply not a lever.
• Rewriting for "AI-friendly" style or chunking pages. Google says you don't need to write in a specific way for generative AI search, and says there is no ideal page length. C-SEO Bench found most rewriting methods ineffective and often harmful.
• Chasing an E-E-A-T score or domain authority. Google says E-E-A-T isn't a specific ranking factor. Domain Authority is a Moz metric no engine computes.
• Quoting "up to 40% more visibility". That figure came from a lab engine with sources already loaded into context.
• Reporting an AI rank. No engine exposes position data, so the number cannot be checked.
• Mass-publishing AI pages. As Otterly's test showed, the Google removal can arrive without warning.